u. brinck detection of inﬂammation- and a neoplasia ... · oooooooooooooooooooo detection of...

© 1998 S. KargerAG, Basel0001–5180/98/1614–0219$15.00/0

Fax¤+¤41 61 306 12 34E-Mail [email protected] Accessible online at:www.karger.com http://BioMedNet.com/karger

U. Brinck a

M. Korabiowska b

R. Bosbach a

H.-J. Gabius c

a Department of Gastroenterologic Pathologyand

b Center of Pathology, Faculty of Medicine,University of Göttingen, and

c Institute of Physiological Chemistry,Faculty of Veterinary Medicine,Ludwig-Maximilians-University,Munich, Germany

Acta Anat 1998;161:219–233

Ulrich Brinck, MDDepartment of Gastroenterologic PathologyFaculty of Medicine, University of GöttingenRobert-Koch-Strasse 40, D–37075 Göttingen (Germany)Tel. +49 (551) 39 8677, Fax +49 (551) 39 8627

oooooooooooooooooooooooooooooooooooooooooooo

Key WordsInflammationAdenomaLarge intestineLectinGalectinNeoglycoprotein

ooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

AbstractCommonly, plant and invertebrate lectins are accepted glycohistochemical tools for the anal-ysis of normal and altered structures of glycans in histology and pathology. Mammalianlectins and neoglycoproteins are recent additions to this panel for the detection of lectin-reac-tive carbohydrate epitopes and glycoligand-binding sites. The binding profiles of these threetypes of probes were comparatively analyzed in normal, inflamed and neoplastic large intes-tine. In normal colonic mucosa the intracellular distribution of glycoconjugates and carbo-hydrate ligand-binding sites in enterocytes reveals a differential binding of lectins with dif-ferent specificity and of neoglycoproteins to the Golgi apparatus, the rough and smoothendoplasmic reticulum and the apical cell surface. The accessible glycoligand-binding sitesand the lectin-reactive carbohydrate epitopes detected by galectin-1 show the same pattern ofintracellular location excluding the apical cell surface. Lectin-reactive carbohydrate epitopesdetected by plant lectins of identical monosaccharide specificity as the endogenous lectin[Ricinus communis agglutinin-I (RCA-I), Viscum album agglutinin (VAA)], however, clearlydiffer with respect to their intracellular distribution. Maturation-associated differences andheterogeneity in glycohistochemical properties of epithelial cells and non-epithelial cells(macrophages, dendritic cells, lymphocytes) are found. Dissimilarities in the fine structuralligand recognition of lectins with nominal specificity to the same monosaccharide have beendemonstrated for the galactoside-specific lectins RCA-I, VAA and galectin-1 as well as theN-acetylgalactosamine (GalNAc)-specific lectins Dolichos biflorus agglutinin (DBA), soy-bean agglutinin (SBA) and Helix pomatia agglutinin in normal mucosa and in acute appen-dicitis. Acute inflammation of the intestinal mucosa found in acute phlegmonous appendicitisis associated with selective changes of glycosylation of mucin in goblet cells mainly of lowerand middle crypt segments resulting in an increase of DBA- and SBA-binding sites in the gob-let cell population. Appendicitis causes no detectable alteration of neoglycoprotein binding. Incontrast, tumorigenesis of colonic adenoma is characterized by increases in lectin-reactivegalactose (Gal; Gal-β1,¤3-GalNAc), fucose and N-acetylglucosamine moieties and by en-hanced presentation of respective carbohydrate ligand-binding capacity. This work revealsthat endogenous lectins and neoglycoproteins are valuable glycohistochemical tools supple-menting the well-known analytic capacities of plant lectins in the fields of gastrointestinalanatomy and gastroenteropathology.oooooooooooooooooooo

Detection of Inflammation- andNeoplasia-Associated Alterationsin Human Large IntestineUsing Plant/Invertebrate Lectins,Galectin-1 and Neoglycoproteins

Abbreviations used in this paper:Con A¤=¤Concanavalin A; DBA¤=¤Dolichos biflorus agglu-tinin; Fuc¤=¤fucose; Gal¤=¤galactose; GalNAc¤=¤N-acetyl-galactosamine; GALT¤=¤gut-associated lymphoid tissue;Glc¤=¤glucose; GlcNAc¤=¤N-acetylglucosamine;HPA¤=¤Helix pomatia agglutinin; Lac¤=¤lactose; Mal¤=¤malt-ose; Man¤=¤mannose; RCA-I¤=¤Ricinus communis agglu-tinin-I; Rham¤=¤rhamnose; SBA¤=¤soybean agglutinin;UEA-I¤=¤Ulex europaeus agglutinin; VAA¤=¤Viscum albumagglutinin.

Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM

Introduction

The primary function of the large intestine is to absorbthe fluid remaining in the luminal contents during diges-tion. As the fecal mass develops in the colon, secretionsfrom the large population of goblet cells in the mucosa pro-vide the lubrication necessary to prevent mucosal damageand infection as the feces pass through the colon. The gut-associated lymphoid tissue (GALT) of the large intestinecooperates with the mucosal epithelium in the defenseagainst potentially pathogenic microorganisms in the gutlumen [Filipe, 1979; Podolsky, 1989].

Glycoproteins are functionally involved in these ‘non-specific’ and immunological defense mechanisms, sincethey are important constituents of the mucin produced byintestinal goblet cells and also take part in the binding anduptake of intraluminal antigenic material and potentiallypathogenic microorganisms [Neutra et al., 1987]. Their gly-can composition is not a constant feature and is subject tomodulation. Recent evidence suggests that glycomoiety-binding proteins (lectins) used as histochemical tools revealquantitative variations in the expression of binding sites innormal tissue of the large intestine in relation to the statusof differentiation and activation of epithelial and nonepi-thelial cell types (macrophages, dendritic cells, lymphoidcells) involved in defense mechanisms [Brinck et al., 1995,1996]. In addition, disease-associated alterations of lectin-reactive carbohydrate epitopes and glycoligand-bindingsites of some of these cell types may occur during inflam-matory reactions and neoplastic transformation.

To provide a further example of the illustration of thetechnical feasibility to closely monitor glycohistochemicalchanges and of the ensuing results of the application ofexogenous and endogenous lectins and neoglycoproteins,we will present and discuss the detection of lectin-reactivecarbohydrate epitopes and glycoligand-binding sites in dif-ferent segments of the normal human large intestine andin relation to inflammatory and neoplastic disease. Glyco-histochemical findings will be summarized at first for theproximal large intestine, which includes the vermiform ap-pendix, cecum and the proximal part of the colon and thencompared to the findings in the normal distal large intestine(colon sigmoideum and rectum). Inflammation-related al-terations are exemplified with respect to acute phlegmo-nous appendicitis. Histochemical findings in neoplasticdisease focus on adenoma of the distal colon to show earlyalterations within the adenoma-carcinoma sequence.

Theoretical Basis for Selection ofGlycohistochemical Tools

The glycan part of glycoconjugates has intrigued re-searchers in various fields for decades [Sharon, 1998]. Tra-ditionally, plant and invertebrate lectins are used to deter-mine whether distinct lectin-reactive carbohydrate epitopesare present [Danguy et al., 1994]. The chapters of Danguyet al. [1998], Mann and Waterman [1998], Plendl andSinowatz [1998] and Zschäbitz [1998] further illustrate thisapplication. Notably, the chapter of Rüdiger [1998] answersthe question concerning in vivo functions of plant lectinswhich has been elusive for decades despite the popularityof the agglutins as laboratory instruments. Biochemicalwork over the last two decades has proven that lectins arealso a functionally crucial part of mammalian cells [Gabius,1987, 1997a, b; for a collection of recent reviews, seeGabius and Gabius, 1997; Kaltner and Stierstorfer, 1998;Zanetta, 1998]. To go beyond the mentioned monitoringwith tools of non-mammalian origin, it is reasonable to ex-tend such a panel by employing lectins isolated from mam-malian tissue [Gabius et al., 1993]. To document this ap-plication, a mammalian galectin (galectin-1) was includedin a selection of exogenous agglutinins with specificitiesto common constituents of cellular glycoconjugates. Com-pared to the direct analysis of glycoconjugate structure, out-lined by Geyer and Geyer [1998] in this issue, lectin histo-

220 Acta Anat 1998;161:219–233 Brinck/Korabiowska/Bosbach/Gabius

Table 1. Glycohistochemical tools applied in histochemicalanalyses of the large intestine [Brinck et al., 1995, 1996]

¤Lectins Source Specificity Specificity ofof of correspondinglectin lectins carbohydrate ligand-

exposing neoglycoproteins

Con A plant α-D-Man α-D-Manα-D-Glc Mal

UEA-I plant α-L-Fuc α-L-FucDBA plant α-D-GalNAc –SBA plant D-GalNAc β-D-GalNAcHPA invertebrate D-GalNAc β-D-GalNAcRCA-I plant β-D-Gal LacVAA plant D-Gal LacGalectin-1 mammalian D-Gal Lac– – – α-L-Rham– – – β-D-GlcNAc

Man¤=¤Mannose; Glc¤=¤glucose; Fuc¤=¤fucose; GalNAc¤=¤N-acetyl-galactosamine; Gal¤=¤galactose; Mal¤=¤maltose; Lac¤=¤lactose; Rham¤=rhamnose; GlcNAc¤=¤N-acetylglucosamine.

Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM

chemistry provides an insight into the presence of distinctglycan chain constituents in situ. Interestingly, this lectinis involved in growth-controlling mechanisms in culturedhuman neuroblastoma cells [Kopitz et al., 1998], providingan example of lectin-dependent, physiologically relevantsignaling [Villalobo and Gabius, 1998].

As there is a growing notion to accept the idea thatglycans are excellent information-storing elements [Laine,1997], their ligand potential can be exploited for the detec-tion of receptor sites, prompting the synthesis of neoglyco-conjugates. Consequently, carbohydrate structures werechemically conjugated to a histochemically inert, labeledcarrier, thus establishing a neoglycoconjugate and therebymaking it possible to conveniently monitor the potentialligand properties of such epitopes [Gabius, 1988; Lee andLee, 1994; Bovin and Gabius, 1995; Danguy et al., 1998].Combined chemoenzymatic synthesis has recently evengained access to neoglycoconjugates with complex N-gly-cans as ligand part [André et al., 1997]. These tools wereused to determine the expression of accessible carbohy-drate-binding sites which were not negatively affected bythe processing of tissue specimen. For example, in breastand lung cancer these tools have described tumor type-associated differences in the capacity to bind sugar ligandsbased on lectin expression [Gabius et al., 1986, 1988; Kay-ser et al., 1989; Kayser and Gabius, 1997].

Owing to the documented presence of lectins in coloncancer, initially demonstrated by affinity chromatography[Gabius et al., 1985], and the modulation of their expres-sion by chemical agents inducing differentiation in culture

[Gabius et al., 1990], it is of interest to perform a compara-tive analysis with lectins and neoglycoproteins to assessboth sides of a potential recognition system. The types ofcarbohydrate ligands were deliberately chosen to be com-plementary to the specificities of the applied agglutinins toallow a comparison (table 1).

Normal Large Intestine

Glycoconjugate Expression in Different CellularConstituents of Enterocytes (table 2)In the epithelium of normal human proximal large intes-

tine, cellular components of enterocytes differed in theircapacity to present lectin-reactive epitopes. The apical sur-face of enterocytes in paraffin-embedded sections presentedbinding sites for Ricinus communis agglutinin-I (RCA-I)

221Glycohistochemical Detection ofInflammation- and Neoplasia-AssociatedAlterations in Large Intestine

Acta Anat 1998;161:219–233

Table 2. Binding of lectins to epithelium of normal proximal large intestine

¤Location of lectin binding Con A UEA-I DBA SBA HPA VAA RCA-I Galectin-1

Surface enterocytesApical cell surface –/0 +++/4 +/4 ++/3 +++/2 +/3 +++/4 –/0Subapical cytoplasm +++/4 +++/3 ++/3 +++/3 +++/3 ++/1 +++/4 +++/2Supranuclear cytoplasm +++/4 +++/4 ++/3 +++/3 +++/3 ++/2 +++/4 +++/2Pararetronuclear cytoplasm +++/4 ++/3 (+)/2 –/0 (+)/3 –/0 (+)/1 +++/2Crypt enterocytesSubapical cytoplasm +++/4 +++/3 ++/3 +/2 +++/3 (+)/2 +++/3 +++/2Supranuclear cytoplasm +++/4 +++/3 ++/3 +/2 +++/3 (+)/3 +++/3 +++/2Pararetronuclear cytoplasm +++/4 ++/3 (+)/1 –/0 (+)/3 –/0 (+)/1 +++/2Goblet cellsIntracellular mucus –/0 +/2 (+)/3 (+)/3 ++/3 (+)/2 +++/2 (+)/1Luminal mucus –/0 +++/3 (+)/3 +/3 ++/3 (+)/2 +++/4 –/0

The percentage of positive structures (cellular subsites, cells, mucus) is grouped into the categories: –¤=¤0%; (+)¤=¤0–20%: +¤=¤20–40%;++¤=¤40–60%; +++¤=¤60–100%. The intensity of staining is grouped into the categories: 0¤=¤no staining; 1¤=¤weak, but significant staining;2¤=¤medium staining; 3¤=¤strong staining; 4¤=¤very strong staining [from Brinck et al., 1995, with modifications].

Fig. 1. Light micrographs of normal human appendiceal surfaceepithelium (A–G) after stepwise application of biotinylated lectins,namely the galactoside-specific lectin from mistletoe (VAA; B), themammalian β-galactoside-specific lectin galectin-1 (C), Con A (D),SBA (E), DBA (F) and galectin-1 in the presence of 0.2 M lactose fordemonstration of inhibition (G), as well as of biotinylated lactosylatedalbumin (A), ABC reagents, the chromogen 3-amino-9-ethylcarba-zole and hematoxylin counterstaining. ×¤2,500.

Fig. 2. Light micrographs of normal human appendiceal mucosaafter stepwise application of biotinylated lectins, namely SBA (A) andHPA (B), ABC reagents, the chromogen 3-amino-9-ethylcarbazoleand hematoxylin counterstaining. ×¤330.

Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM


1(For legend, see p. 221.)

Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM


Acta Anat 1998;161:219–233

2(For legend, see p. 221.)

Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM

on all cells, binding sites for Ulex europaeus agglutinin(UEA-I), Dolichos biflorus agglutinin (DBA), soybean ag-glutinin (SBA), Helix pomatia agglutinin (HPA) and Viscumalbum agglutinin (VAA) heterogeneously and no bindingsites for concanavalin A (Con A) and galectin-1. This isillustrated exemplarily for VAA, Con A and galectin-1 infigure 1B–D. Presence of SBA- and VAA-binding sites wasrestricted to the supranuclear and subapical cell regions(fig. 1E). UEA-I-, DBA-, HPA- and RCA-I-binding siteswere located both in the supranuclear and subapical cellregions and, quantitatively different, in the basal cell re-gion. Reactive glycoconjugates for all these lectins (SBA,VAA, RCA-I, UEA-I, DBA, HPA) were present focaly inthe cells, thus suggesting binding primarily to organelles,most probably of the Golgi apparatus and the smooth endo-plasmic reticulum in the subapical cell region [Lee, 1987].This is demonstrated exemplarily for the agglutinins SBAand DPA in figure 1E and F. Con A and galectin-1 werebound to glycoligands in the cytoplasm of the supranuclear,subapical and basal cell regions (fig.¤1C, D). The preferenceof Con A for cytoplasmic staining was attributable to thepresence of oligomannose-like structures in the roughendoplasmic reticulum [Laurila et al., 1978; Fuhrman andBereiter-Hahn, 1984].

Intracellular Distribution of CarbohydrateLigand-Binding Sites within Enterocytes (table 3)The epithelium of the intestinal mucosa not only pre-

sented lectin-binding sites, but also harbored the capacity tobind carbohydrate structures, as shown by labeled glyco-ligand-exposing neoglycoproteins (fig.¤1A). Intracellulardistribution of carbohydrate ligand-binding sites within en-terocytes of appendiceal and colonic mucosa resembled thebinding pattern of the mammalian galectin-1 more closelythan that of any other above-mentioned lectins. This simi-larity of cellular binding patterns is illustrated for lactose-binding sites and the mammalian galectin (fig.¤1A, C). Asthe presence of β-galactoside-binding sites intimates a po-tential interaction between the two types of epitopes, local-ized by an endogenous lectin and a lectin-detecting marker,functional implications for carbohydrate-lectin interactionsin situ are encouraged. Binding of neoglycoproteins ex-cluded the apical cell surface with the striated border, andthe secretory vesicles of goblet cells. This appearance canbe interpreted to reflect a masking of specific receptors atcertain locations by an abundant presence of glycoconju-gates. The potential of mucins to associate strongly withendogenous lectins has for example been underscored bythe demonstration that galectins can bind colonic mucins[Bresalier et al., 1996; Wasano and Hirakawa, 1997].

Indications for Maturation-Associated Differences inGlycohistochemical Properties of Epithelial CellsRemarkable quantitative differences were found for

lectin binding between the enterocytes of crypts and of thesurface epithelium. Enterocytes of the surface epitheliumare derived from the crypt epithelium [Wolf and Bye, 1984]and establish a functionally specialized cover called thefollicle-associated epithelium [Bockman, 1983]. Therefore,it is reasonable to suggest that these differences are relatedto the status of differentiation. Apparently, these differencesmainly affected the presentation of binding sites of SBA(fig.¤2A) and VAA, which appeared more frequently inenterocytes of the follicle-associated epithelium than inenterocytes of the crypt epithelium, and of HPA which didnot bind to the basolateral membranes of the follicle-asso-ciated epithelium.

Maturation-associated differences of expression of lec-tin-reactive carbohydrate epitopes were also found withinthe goblet cell population. Within this cell type, the pres-ence of N-acetylgalactosamine (GalNAc) residues detectedby DBA or SBA was restricted to the upper third of thecrypts and the surface epithelium (fig.¤2A). The intensityof binding of carrier-immobilized carbohydrate ligandssuch as β-galactose, β-GalNAc, β-N-acetylglucosamine andα-mannose to epithelial cells of normal mucosa of the largeintestine is apparently correlated with the status of matura-tion of the cells. There was an increased staining intensityextending from the bottom of the crypts to the surfaceepithelium. Such a pattern of labeling was also found forSBA (fig.¤2A) and RCA-I (fig.¤4A) in the proximal largeintestine as well as for Con A and galectin-1 in enterocytesof the distal large intestine, making it possible to assumethat the phenomenological result might have a physiologi-cal basis [Brinck et al., 1995].

Heterogeneity of Glycohistochemical Properties ofEnterocytes and Goblet CellsHeterogeneity of epithelial cells within single crypts of

the colonic mucosa with respect to the expression of plantand invertebrate lectin-reactive carbohydrate epitopes wasobvious and contrasted with a rather homogeneous distri-bution of accessible glycoligand-binding sites. Goblet cellmucus displayed heterogeneous distributions of lectin-binding sites: binding sites for Con A were not detected ingoblet cells, for DBA, SBA, VAA and galectin-1 in lessthan 20%, for UEA-I in 20–40%, for HPA in 40–60% andfor RCA-I in 60–100% of the goblet cells.

The affinity of the apical surface of enterocytes includ-ing M cells differed notably among the lectins studies, asindicated by different staining intensities, heterogeneity of


Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM

binding to enterocytes or lack of apical surface binding. Theapical surface of enterocytes is formed by microvilli withmucus intimately associated with it [Forstner et al., 1973;Etzler, 1979]. Since lectin-binding sites at the apical surfaceof intestinal epithelial cells can be used as adhesion sites bybacteria, protozoa, or viruses, lectin-binding sites in this lo-cation may play a role in intestinal infection [Gitler et al.,1985; Sharon, 1987; Doyle and Slifkin, 1994]. Galactoseresidues visualized on the apical surface of M cells and otherenterocytes might for example provide suitable adhesionsites for Entamoeba histolytica with specificity to its surfacelectins [Petri et al., 1987]. In this respect heterogeneity of theglycohistochemical features of enterocytes may be of inter-est. Based on the observed correlation of the presence oflectin-binding sites of DBA, SBA and VAA at the apical sur-face of enterocytes with the presence of equivalent bindingsites in the supranuclear cell region, it can be assumed that

the heterogeneity of respective lectin-binding sites at the api-cal surface may be due to differences in the synthetic ma-chinery of enterocytes rather than to accidental local varia-tions in mucin binding in the tissue specimen. As reviewedby Brockhausen et al. [1998] and Hakomori [1998] in thisissue, the cascade of glycosylation steps can be subject to al-terations correlating for example with disease.

Glycohistochemical Subtyping of Macrophages andDendritic Cells in the GALT (tables 3, 4)In Peyer’s plaques of normal vermiform appendix the

ligand density for lectins can differ for cells with respect tocell type and location of cells. A panel of lectins (UEA-I,DBA, SBA, HPA, RCA-I, galectin-1) is useful for distin-guishing three subtypes of macrophages in the GALT ofthe human vermiform appendix on the basis of quantitativedifferences in glycosylation. They comprise cells near the


Acta Anat 1998;161:219–233

Table 3. Binding of carrier-immobilized carbohydrate ligands to epithelium of the normal large intestine and to Peyer’s patches of normalvermiform appendix

¤Site of neoglycoprotein binding Lac β-GalNAc β-GlcNAc α-Man α-L-Fuc Mal α-L-Rham

Epithelial cellsSurface enterocytes +++/2 +++/2 +++/3 +++/3 +++/4 +++/1 +++/2Crypt enterocytes +++/1 +++/1 +++/2 +++/2 +++/2 +++/1 +++/1Follicle-associated epitheliumLymphocytes –/0 –/0 –/0 (+)/2 (+)/2 –/0 –/0Macrophages –/0 –/0 –/0 (+)/2 (+)/2 –/0 –/0DomeLymphocytes –/0 –/0 –/0 (+)/2 (+)/2 –/0 –/0Plasma cells –/0 –/0 –/0 ++/3 ++/3 –/0 –/0Macrophages –/0 –/0 –/0 (+)/2 (+)/2 –/0 –/0Intercryptal regionLymphocytes –/0 –/0 –/0 (+)/2 (+)/2 –/0 –/0Plasma cells –/0 –/0 –/0 ++/3 ++/3 –/0 –/0Macrophages –/0 –/0 –/0 (+)/2 (+)/2 –/0 –/0Mantle zoneLymphocytes –/0 –/0 –/0 +/2 +/2 –/0 –/0Germinal centerLymphoid cells (inner two thirds) –/0 –/0 –/0 –/0 –/0 –/0 –/0Lymphoid cells (outer third) –/0 –/0 –/0 –/0 –/0 –/0 –/0Macrophages –/0 –/0 –/0 –/0 –/0 –/0 –/0Dendritic reticulum cells –/0 –/0 –/0 –/0 –/0 –/0 –/0T regionLymphocytes –/0 –/0 –/0 +/2 +/2 –/0 –/0Macrophages –/0 –/0 –/0 –/0 –/0 –/0 –/0Interdigitating reticulum cells –/0 –/0 –/0 –/0 –/0 –/0 –/0

The percentage of positive cells is grouped into the categories: –¤=¤0%; (+)¤=¤0–20%: +¤=¤20–40%; ++¤=¤40–60%; +++¤=¤60–100%. The inten-sity of staining is grouped into seven categories of increasing intensity, ranging from 1 (weak, but significant staining) to 7 (strong staining).Lac¤=¤Lactose; GalNAc¤=¤N-acetylgalactosamine; GlcNAc¤=¤N-acetylglucosamine; Man¤=¤mannose; Fuc¤=¤fucose; Mal¤=¤maltose; Rham¤=¤rham-nose [from Brinck et al., 1996, with modifications].

Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM

lumen, that is macrophages of the follicle-associated epi-thelium and subepithelial macrophages of the dome andlamina propria, tingible body macrophages of the germinalcenters and macrophages of the T region. Macrophagesnear the gut lumen consistently expressed the broadestspectrum of lectin-reactive carbohydrate structures. UEA-Iand DBA were never bound to any of the tingible bodymacrophages of the germinal centers or the T region. Thedifferential expression of lectin-reactive carbohydrate epi-topes in tingible body macrophages of germinal centers isillustrated for the binding sites of the GalNAc-specificlectins DBA, SBA and HPA as well as the mannose- andglucose-specific lectin Con A in figure 3. UEA-I and DBAappeared to bind preferentially to lysosomal structures.Since it is common knowledge that foreign antigenicmacromolecules taken up from the gut lumen are also trans-ported to macrophages in germinal centers [von Rosen etal., 1981], this reduction in lectin binding may be attribut-able to a degradation of phagocytized material in deeperportions of Peyer’s plaques and/or to altered (physicochem-ical) properties of lysosomal membranes.

A second location-dependent peculiarity was observedin macrophages which were either located directly beneaththe basal membrane of the follicle-associated epitheliumor in the follicle-associated ephithelium itself. These mac-rophages can bind VAA at the cell surface which is incontact with the basal membrane, lymphocytes or M cells.Surface binding of VAA was not observed in macrophagesof other regions. It is likely, but not proven beyond doubtthat this observed property is associated with macrophagestimulation in this special environment, for example anti-


Table 4. Binding of lectins to subregions of appendiceal Peyer’s patches

¤Location of lectin binding Con A UEA-I DBA SBA HPA VAA RCA-I Galectin-1

Follicle-associated epitheliumLymphocytes +/2 –/0 –/0 –/0 –/0 (+)/3 +++/3 (+)/1Macrophages ++/3 ++/3 ++/2 ++/3 ++/2 ++/3 +++/3 +/1DomeLymphocytes +++/2 –/0 –/0 –/0 –/0 (+)/1 +++/4 ++/2Plasma cells +++/3 –/0 –/0 –/0 +/1 –/0 +++/4 ++/2Macrophages +++/3 ++/2 ++/2 +++/3 +++/3 ++/3 +++/4 ++/2Intercryptal regionLymphocytes ++/1 –/0 –/0 –/0 –/0 (+)/1 +++/4 +/2Plasma cells +++/3 –/0 –/0 –/0 +++/2 –/0 +++/4 ++/2Macrophages +++/3 ++/2 ++/2 +++/3 +++/3 +/3 +++/4 ++/2Mantle zoneLymphocytes ++/1 –/0 –/0 –/0 –/0 (+)/1 +++/4 +/1Germinal centerLymphoid cells (inner two thirds) +++/2 –/0 –/0 –/0 –/0 (+)/1 +++/4 +/2Lymphoid cells (outer third) +++/1 –/0 –/0 –/0 –/0 (+)/1 +++/3 +/2Macrophages +++/4 –/0 –/0 +++/3 ++/2 +++/4 +++/4 ++/2Dendritic reticulum cells –/0 –/0 –/0 –/0 –/0 –/0 –/0 –/0T regionLymphocytes ++/1 –/0 –/0 –/0 –/0 ++/2 ++/2 –/0Macrophages +++/3 –/0 –/0 +/3 +/3 +/1 +++/3 –/0Interdigitating reticulum cells +++/3 –/0 –/0 +/2 ++/2 (+)/2 +++/2 –/0

The percentage of positive cells is grouped into the categories: –¤=¤0%; (+)¤=¤0–20%: +¤=¤20–40%; ++¤=¤40–60%; +++¤=¤60–100%. The inten-sity of staining is grouped into the categories: 0¤=¤no staining; 1¤=¤weak, but significant staining; 2¤=¤medium staining; 3¤=¤strong staining; 4¤=¤verystrong staining [from Brinck et al., 1996, with modifications].

Fig. 3. Light micrographs of normal human appendiceal germi-nal center after stepwise application of biotinylated lectins, namelyDBA (A), SBA (B), HPA (C) and Con A (D), ABC reagents, the chro-mogen 3-amino-9-ethylcarbazole and hematoxylin counterstaining.×¤380.

Fig. 4. Light micrographs of normal human appendiceal mucosaafter stepwise application of biotinylated lectins, namely RCA-I (A)and the mammalian β-galactoside-specific lectin galectin-1 (B), ABCreagents the chromogen 3-amino-9-ethylcarbazole and hematoxylincounterstaining. ×¤330.

Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM


Acta Anat 1998;161:219–233

(For legends, see p. 226.)4

3

Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM

gen exclusion by M cells. Lectin binding on the surface ofmacrophages is of special relevance for phagocytosis, be-cause these binding sites may be involved in adsorption ofextracellular macromolecules (with lectin properties), forexample bacterial surface lectins [Ofek and Sharon, 1988].In addition, glycoconjugates on the surface of macrophagesin the follicle-associated epithelium or in a subepithelialposition could interact with lectins of other cells likelymphocytes [Gabius, 1987, 1997a; Grillon et al., 1990;Abramenko et al., 1992; Sharma et al., 1992] which may beespecially relevant at the conditions of activation. Notably,galectin ligands can under such conditions be downregu-lated, ascertaining the possibility for regulation on the levelof binding partners for distinct tissue lectins [Smetana etal., 1998].

Marked differences were observed between the lectin-binding characteristics of dendritic cells in germinal centersand in the T region. Binding of lectins to interdigitatingcells in the T region was similar to binding to macrophages.This observation corroborates the view that interdigitatingcells are derived from the mononuclear phagocyte system[Radzun et al., 1984]. Our finding that follicular dendriticreticulum cells did not share lectin-binding characteristicsof macrophages and of interdigitating cells is consistentwith the notion that they have a different cellular origin,perhaps originating from perivascular tissue [Beranek andMasseyeff, 1986].

Glycohistochemistry of the GALT (tables 3, 4)Lymphocytes in all anatomical subsites of the GALT,

centrocytes, centroblasts and plasma cells all had commonbinding sites for Con A and RCA-I. This is illustratedexemplarily for RCA-reactive carbohydrate epitopes infigure 4A. Notably, a small number of lymphocytes, mostlyin the T region but also in B-cell-rich areas, expressedintranuclear binding sites for fucose and mannose residues,strongly suggesting a role of nuclear lectins and glycopro-teins, as discussed by Hubert et al. [1989]. Intraepitheliallymphocytes and lymphatic cells of the T region differedfrom lymphocytes in other regions by a more frequentexpression of VAA-binding sites.

Dissimilarities in the Fine Structural LigandRecognition of Lectins with Nominal Specificity to theSame Monosaccharide in Enterocytes,Goblet Cells and MacrophagesHistorically, lectins have been grouped according to

their monosaccharide specificity. This classification system,for example based on inhibition of hemagglutination,should not delude one into concluding that the actual bind-

ing partners of lectins of identical monosaccharide speci-ficity are also identical. In any case, the subtleties of ligandbinding must be thoroughly probed to accurately determinethe fine specificity. To address the complexities presentedby this task, either a battery of oligosaccharides is testedor the structure of the recognition structure is analyzedby techniques, outlined by Geyer and Geyer [1998] in thisissue. To underscore the fact that this problem should not atall be viewed as incidental to histochemical studies whenaspiring to document an analysis of glycan expression, it ispointed out that plant and animal lectins with reactivity togalactose differ notably with respect to the fine specificityregarding sequence extensions beyond the terminal mono-saccharide [Lee et al., 1992, 1994; Galanina et al., 1997;Kaltner et al., 1997]. As proven recently by von der Liethet al. [1998], disaccharides may even undergo differentialconformer selection, reflecting disparities in the architec-ture of the binding sites. Overall, the shared specificity togalactose does not guarantee a comparable profile of bind-ing to subpopulations of cellular galactose-containing glyco-conjugates. This conclusion emphasizes the need to employthe tissue lectin for any functional correlations. The dif-ferential fine specificity of the galactoside-specific lectinsVAA, RCA-I and galectin-1 for cellular constituents in gly-cohistochemical analysis is illustrated in figures 1B, C and4A, B. With respect to the subcellular staining pattern, it isinteresting that only the mammalian lectin exhibited nu-clear staining. Such a result has already been detected in theinitial immunohistochemical study with respect to tumorpathology [Gabius et al., 1986]. It is relevant to mentionthat galectin-1 has a role in pre-mRNA splicing [Vyakar-nam et al., 1997]. As a consequence of different receptorproperties, RCA-I and VAA stained macrophages and lym-phocytes fairly well in contrast to the tissue lectin.

Corroborating the observations for galactoside-bindinglectins, the GalNAc-specific lectins SBA, DBA and HPAsimilarly revealed notable differences in their binding pat-tern to certain cell types, namely enterocytes, goblet cells,macrophages and plasma cells, reflecting measurable levelsof dissimilarities in the fine-structural ligand recognition.This is quite well illustrated for epithelial and endothelialcells of the intestinal mucosa as well as for macrophages of


Fig. 5. Light micrographs of normal human appendiceal crypts(A, C) and appendiceal crypts in acute appendicitis (B) as well asof adenoma of the large intestine (D) after stepwise application ofbiotinylated SBA (A, B) and lactosylated albumin (C, D), ABCreagents, the chromogen 3-amino-9-ethylcarbazole and hematoxylincounterstainin. (A, B ×¤330. C ×¤800. D ×¤500.

Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM


Acta Anat 1998;161:219–233

Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM

the germinal centers which were labeled by HPA and SBA,whereas DBA did not bind to these cells (fig. 2A, B, 3A–C).

In this context, one should bear in mind that evenclosely related lectins of the same monosaccharide speci-ficity from the organism may exhibit differential bindingcapacity to homologous lymphocyte populations, for exam-ple the two avian galectins [Schneller et al., 1995]. The spa-tial organization of ligands may influence their reactivitiesbesides their sequences and conformations, as expertly dis-cussed for selectins [Varki, 1994]. To be able to dock oncomplementary sites, the presentation of binding pockets,as provided by crystallographic analysis, will have an im-pact on the level of affinity in situ. It is the indisputablestrength of tailor-made neoglycoconjugates to correlate thespatial factor of optimal geometric ligand arrangement withaffinity [Lee and Lee, 1994, 1997; André et al., 1997].

Comparative Glycohistochemical Analysis of theProximal and Distal Large IntestineA comparative study of lectin binding to goblet cell

mucin in another region of the large intestine, namely therectosigmoid, demonstrated that DBA, SBA and galectin-1primarily bound to the distal colon, while UEA-I and VAAlabeling was selectively found in goblet cell mucin of theproximal large intestine. These spatial differences must betaken into consideration in glycohistochemical studies ofthe diseased colon, in which altered tissue will be comparedwith control specimens of the mucosa from the same intes-tinal region of normal individuals.

Glycohistochemical Alterations inAcute Inflammation Demonstrated inPhlegmonous Appendicitis

The percentage of goblet cells expressing DBA- andSBA-binding sites in mucus globules was found to be about4 times higher in appendicitis than in the normal appendix(fig.¤5A, B). These results demonstrate that the expressionof lectin-binding sites in appendiceal goblet mucin isspecifically altered in appendicitis, indicating that there areselective changes of glycosylation of mucin in goblet cellsmainly of the lower and middle crypt segment.

Similar changes of glycosylation of goblet cell mucus(increase of extent of DBA and SBA binding, but not ofHPA binding to goblet cell mucus) had been assessed ininflammatory bowel disease [Yoshioka et al., 1989]. Thus,appendicitis can be regarded as an appropriate model sys-tem, demonstrating that changes of mucus glycosylationcan be related to acute inflammation rather than a mixture

of chronic and acute inflammatory processes as in inflam-matory bowel disease. However, the qualitatively identicalchange of mucus glycosylation (in terms of lectin binding)as in inflammatory bowel disease is not necessarily relatedto malignant change, as has been controversely discussedfor inflammatory bowel disease [Ahnen et al., 1987]. Al-though the precise functional implications of the observedalterations are at present not obvious, as in a general con-text further discussed by Brockhausen et al. [1998] andHakomori [1998] in this issue, such changes clearly reflecta potentially pertinent impact of the regulation of glycosy-lation, warranting further studies.

Glycohistochemical Alterations inColonic Adenoma

Numerous studies have described alterations of glyco-conjugates associated with tumorigenesis of colonic ade-noma [Boland et al., 1982; Rhodes et al., 1986; Campoet al., 1988; Ho et al., 1988; Lee, 1988; Ota et al., 1988;McGarrity et al., 1989; Orntoft et al., 1991; Dall’Olio andTrere, 1993; Fucci et al., 1993; Jass et al., 1993]. The mostfrequently detected changes of lectin binding in the ade-noma were an increase in the receptivity for peanut agglu-tinin/Amaranthin [Boland et al., 1982; Rhodes et al., 1986;Campo et al., 1988; Lee, 1988; Ota et al., 1988; McGarrityet al., 1989; Orntoft et al., 1991; Sata et al., 1992; Fucciet al., 1993], UEA-I [Rhodes et al., 1986; Ota et al., 1988;McGarrity et al., 1989; Jass et al., 1993] and Griffoniasimplicifolia agglutinin-II [Rhodes et al., 1986; Ota et al.,1988]. As discussed in the preceding paragraph, the appliedtools contribute to the structural analysis of glycans. How-ever, a functional correlation in terms of recognitive inter-play can only be inferred by the detection of suitable recep-tor sites and by the demonstration of ligand properties toendogenous lectins (table 5).

Having synthesized markers to measure the ligand prop-erties of carbohydrate moieties and having purified endoge-nous lectins, it is possible to compare glycohistochemicallyaccessible carbohydrate moieties and equivalent receptorsin tissue material with defined morphological alterations.The binding of neoglycoconjugates presenting blood grouptrisaccharides has already been shown to be of clinicalsignificance. It correlates for A/H epitopes with morpho-metric features in lung and prostate cancer and with sur-vival in patients with lung cancer [Kayser et al., 1994,1995]. Focusing on colon cancer, initial studies with pri-mary and metastatic lesions caution against the view of asimple picture for metastasis formation [Gabius et al.,


Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM

1991; Irimura et al., 1991; Schoeppner et al., 1995]. Withthe manifestation of an adenoma increases in presentationof lectin-reactive Gal (Galβ-1,¤3-GalNAc), fucose and N-acetylglucosamine moieties appear to be associated withenhanced presentation of carbohydrate ligand-binding ca-pacity. In contrast to this quantitative aspect, the qualitativesubcellular binding pattern of neoglycoproteins to adenomacells continues to closely resemble that of epithelial cells ofnormal mucosa (fig.¤5C, D). The apical surface and mucusof secretory vesicles are apparently free of detectable bind-ing sites. As already mentioned, presence of high-affinityligands can render endogenous galectins inaccessible toneoglycoconjugates, among them glycan chains of laminin,carcinoembryonic antigen and lysosome-associated mem-brane glycoproteins 1 and 2 [Ohannesian et al., 1995; Bre-salier et al., 1996]. In summary, the measured quantitativeincrease intimates a potential role in the progression fromadenoma to carcinoma. Similar to the upregulation ofneoglycoprotein-binding sites in chorionepithelioma cells,reported previously [Gabius et al., 1989], the relevanceof these alterations for cell sociology warrants furtherscrutiny.

Conclusion

Application of a panel of plant/invertebrate and endoge-nous mammalian lectins with similar nominal monosaccha-ride specificity as well as of neoglycoproteins with histo-chemically crucial ligand structures probes complementaryaspects of protein-carbohydrate recognition. Monitoring ofnormal, inflamed and neoplastically transformed tissue ofthe large intestine allows a thorough comparison betweenthe binding patterns of exogenous and endogenous lectins

as well as of carrier-immobilized glycoligands to gain func-tionally valid insights such as the colocalization of lectin-reactive carbohydrate epitopes and glycoligand-bindingsites. The endogenous lectin galectin-1 has an apparentlyvery similar pattern of ligand localization as the carrier-immobilized glycoligand lactose. Its binding profile dif-fered from plant and invertebrate lectins with identicalnominal monosaccharide specificity.

Expression of lectin-reactive carbohydrate epitopes andglycoligand-binding sites appears to be related to the mat-urational status of the epithelial cells and allows subtypingof macrophages and dendritic cells with possible functionalimplications. Glycosylation of goblet cell mucus is specif-ically altered in acute inflammation, as demonstrated inphlegmonous appendicitis. Increases in the extent of pre-sentation of glycoligand-binding sites and lectin-reactivecarbohydrate epitopes in colonic adenoma accompanies thetransition from adenoma to carcinoma. Conceptually, theseresults clearly illustrate the power of combined studiesexploiting the target specificities of endogenous lectins andcarbohydrate ligands to infer physiological protein-carbo-hydrate recognition in situ.


Acta Anat 1998;161:219–233

Table 5. Binding of carrier-immobilized carbohydrate ligands to mucosa and adenoma of the large intestine

¤Site of neoglycoprotein binding Lac β-GalNAc β-GlcNAc α-Man α-L-Fuc Mal α-L-Rham

MucosaSurface enterocytes 2 2 3 3 4 1 2Crypt enterocytes 1 1 2 2 2 1 1AdenomaSurface enterocytes 5 4 5 6 5 4 5Crypt enterocytes 6 5 6 7 6 5 6

The intensity of staining reaction is grouped into seven categories of increasing intensity, ranging from 1 (weak, but significant staining)to 7 (strong staining). Lac¤=¤Lactose; GalNAc¤=¤N-acetylgalactosamine; GlcNAc¤=¤N-acetylglucosamine; Man¤=¤mannose; Fuc¤=¤fucose; Mal¤=maltose; Rham¤=¤rhamnose [from Brinck et al., 1996, with modifications].

Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM


Abramenko, I.V., D.F. Gluzman, E.Y. Korchagina,T.V. Zemlyanukhina, N.V. Bovin (1992) Oli-gosaccharide-binding molecules on the sur-face of human hemopoietic and lymphoidcells. FEBS Lett 307: 283–286.

Ahnen, D.J., G.H. Warren, L.J. Greene, J.W.Singleton, W.R. Brown (1987) Search for aspecific marker of mucosal dysplasia inchronic ulcerative colitis. Gastroenterology91: 1356–1355.

André, S., C. Unverzagt, S. Kojima, X. Dong,C. Fink, K. Kayser, H.-J. Gabius (1997)Neoglycoproteins with the synthetic complexbiantennary nonasaccharide or its α2,¤3/α2,¤6-sialylated derivatives: Their preparation, as-sessment of their ligand properties for purifiedlectins, for tumor cells in vitro, and in tissuesections, and their biodistribution in tumor-bearing mice. Bioconjug Chem 8: 845–855.

Beranek, J.T., R. Masseyeff (1986) Commen-tary. Hyperplastic capillaries and their possibleinvolvement in fibrosis. Histopathology 10:543–546.

Bockman, D.E. (1983) Functional histology ofappendix. Arch Histol Jpn 46: 271–292.

Boland, C.R., C.K. Montogomery, Y.S. Kim(1982) A cancer-associated mucin alteration inbenign colonic polyps. Gastroenterology 82:664–672.

Bovin, N.V., H.-J. Gabius (1995) Polymer-immo-bilized carbohydrate ligands: Versatile chem-ical tools for biochemistry and medical sci-ences. Chem Soc Rev 24: 413–421.

Bresalier, R.S., J.C. Byrd, L. Wang, A. Raz (1996)Colon cancer mucin; A new ligand for theβ-galactoside-binding protein galectin-3. Can-cer Res 56: 4354–4357.

Brinck, U., R. Bosbach, M. Korabiowska, A.Schauer, H.-J. Gabius (1995) Lectin-bindingsites in the epithelium of normal human appen-dix vermiformis and in acute appendicits. His-tol Histopathol 10: 61–70.

Brinck, U., R. Rosbach, M. Korabiowska, A.Schauer, H.-J. Gabius (1996) Histochemicalstudy of expression of lectin-reactive carbohy-drate epitopes and glycoligand-binding sites innormal human appendix vermiformis, colonicmucosa, acute appendicitis and colonic ade-noma. Histol Histopathol 11: 919–930.

Brockhausen, I., I. Schutzbach, W. Kuhns (1998)Glycoproteins and their relationship to humandisease. Acta Anat 161: 36–78.

Campo, E., E. Condom, A. Palacin, E. Quesada,A Cardesa (1988) Lectin binding patterns innormal and neoplastic colonic mucosa. A studyin Dolichos biflorus agglutinin, peanut agglu-tinin, and wheat germ agglutinin. Dis ColonRectum 31: 892–899.

Dall’Olio, F., D. Trere (1993) Expression of α 2,¤6-sialylated sugar chains in normal and neo-plastic colon tissues. Detection by digoxigenin-conjugated Sambucus nigra agglutinin. Eur JHistochem 37: 257–265.

Danguy, A., F. Akif, B. Pajak, H.-J. Gabius(1994) Contribution of carbohydrate histochem-istry to glycobiology. Histol Histopathol 9:155–171.

Danguy, A., C. Decaestecker, F. Genten, I. Salmon,R. Kiss (1998) Applications of lectins and neo-glycoconjugages in histology and pathology.Acta Anat 161: 206–218.

Danguy, A., K. Kayser, N.V. Bovin, H.-J. Gabius(1995) The relevance of neoglycoconjugatesfor histology and pathology. Trends GlycosciGlycotechnol 7: 261–275.

Doyle, R.J., M. Slifkin (eds) (1994) Lectin-Mi-croorganism Interactions. New York, Dekker.

Etzler, M.E. (1979) Lectins as probes in studies ofintestinal glycoproteins and glycolipids. Am JClin Nutr 32: 133–138.

Filipe, M.I. (1979) Mucins in the human gastro-intestinal epithelium: A review. Invest CellPathol 2: 195–216.

Forstner, J., N. Taichman, V. Kalnins, G. Forstner(1973) Intestinal goblet cell mucus: Isolationand identification by immunofluorescence ofa goblet cell glycoprotein. J Cell Sci 12:586–602.

Fucci L., A.M. Valentini, M.L. Caruso (1993) Canpeanut agglutinin distinguish between pseudoand true invasion in colonic adenomas? Eur JHistochem 37: 335–344.

Fuhrman, C., J. Breiter-Hahn /1984) Coincidenceof endoplasmic reticulum pattern as visualizedby FITC-Con A-fluorescence and electron mi-croscopy. Histochemistry 80: 153–156.

Gabius, H.-J. (1987) Endogenous lectins in tumorsand the immune system. Cancer Invest 5:39–46.

Gabius, H.-J. (1988) Tumor lectinology: At theintersection of carbohydrate chemistry, bio-chemistry, cell biology and oncology. AngewChem Int Ed 27: 1267–1276.

Gabius, H.-J. (1991) Detection and functions ofmammalian lectins – With emphasis on mem-brane lectins. Biochim Biophys Acta 1071:1–18.

Gabius, H.-J. (1997a) Animal lectins. Eur J Bio-chem 243: 543–576.

Gabius, H.-J. (1997b) Concepts of tumor lectinol-ogy. Cancer Invest 15: 454–464.

Gabius, H.-J., S. Bodanowitz, A. Schauer (1988)Endogenous sugar-binding proteins in humanbreast tissue and benign and malignant breastlesions. Cancer 61: 1125–1131.

Gabius, H.-J., R. Brehler, A. Schauer, F. Cramer(1986) Localization of endogenous lectins innormal human breast, benign breast lesionsand mammary carcinomas. Virchows Arch 52:107–115.

Gabius, H.-J., P.L. Debbage, N. Lang, W. Lange(1989) Malignant and normally developingtrophoblastic cells of human placenta displaydifferent characteristics defined by histochem-ical and biochemical mapping of endogenouslectins. Histochemistry 92: 283–289.

Gabius; H.-J., R. Engelhardt, F. Cramer, R. Bätge,G.A. Nagel (1985) Pattern of endogenous lec-tins in a human epithelial tumor. Cancer Res45: 253–257.

Gabius, H.-J., S. Gabius (1997) Glycosciences:Status and Perspectives. Weinheim, Chapman& Hall.

Gabius, H.-J., S. Gabius, T.V. Zemlyanukhina,N.V. Bovin, U. Brinck, A. Dangui, S.S. Joshi,K. Kayser, J. Schottelius, F. Sinowatz, L.F.Tietze, F. Vidal-Vanaclocha, J.-P. Zanetta(1993) Reverse lectin histochemistry: Designand application of glycoligands for detectionof cell and tissue lectins. Histol Histopathol 8:369–383.

Gabius, H.-J., T. Grote, S. Gabius, U. Brinck, L.F.Tietze (1991) Neoglycoprotein binding tocolorectal tumor cells: Comparison betweenprimary and secondary lesions. Virchows Arch419: 217–222.

Gabius, S., N. Yamazaki, W. Hanewacker, H.-J.Gabius (1990) Regulation of distribution,amount and ligand affinity of sugar receptors inhuman colon carcinoma cells by treatment withsodium butyrate, retinoic acid and phorbolester. Anticancer Res 10: 1005–1012.

Galanina, O.E., H. Kaltner, L.S. Khraltsova, N.V.Bovin, H.-J. Gabius (1997) Further refinementof the description of the ligand-binding charac-teristics for the galactoside-binding mistletoelectin, a plant aggulutinin with immunomodu-latory potency. J Mol Recognit 10: 139–147.

Geyer, H., R. Geyer (1998) Strategies for glyco-conjugate analysis. Acta Anat 161: 18–35.

Gitler, C., M. Mogyoros, E. Calef, I. Rosenberg(1985) Lethal recognition between Entamoebahistolytica and the host tissues. Trans R SocTrop Med Hy 79: 581–586.

Grillon, C., M. Monsigny, C. Kieda (1990) Cellsurface lectins of human granulocytes: Theirexpression is modulated by mononuclear cellsand granulocyte/macrophage colony-stimulat-ing factor. Glycobiology 1: 33–38.

Hakomori, S. (1998) Cancer-associated glycosph-ingolipid antigens: Their structure, organiza-tion and function. Acta Anat 161: 79–90.

Ho, S.B., N.W. Toribara, R.S. Bresalier, J.S. Kim(1988) Biochemical and other markers of co-lon cancer. Gastroenterol Clin North Am 17:811–836.

Hubert, I., A.P. Seve, P. Facy, M. Monsigny (1989)Are nuclear lectins and nuclear glycoproteinsinvolved in the modulation of nuclear func-tions? Cell Differ Dev 27: 69–81.

Irimura, T., Y. Matsushita, R.C. Sutton, D. Carra-lero, D.W. Ohannesian, K.R. Cleary, D.M. Ota,G.L. Nicolson, R. Lotan (1991) Increased con-tent of an endogenous lactose-binding lectin inhuman colorectal carcinoma progressed tometastatic stages. Cancer Res 51: 387–393.

Jass, J.R., Allison, S.M. Stewart, M.R. Lane (1993)Ulex europaeus agglutinin-1 binding in hered-itary bowel cancer. Pathology 25: 114–119.

References

Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM

Kaltner, H., K.S. Lips, G. Reuter, S. Lippert, F.Sinowatz, H.-J. Gabius (1997) Quantitationand histochemical localization of galectin-1and galectin-1-reactive glycoconjugates infetal development of bovine organs. HistolHistopathol 12: 945–960.

Kaltner, H., B. Stierstorfer (1998) Animal lectinsas cell adhesion molecules. Acta Anat 161:162–179.

Kayser, K., N.V. Bovin, E.Y. Korchagina, C.Zeilinger, F.-Y. Zeng, H.-J. Gabius (1994) Cor-relation of expression of binding sites for syn-thetic blood group A-, B-, and H-trisaccharidesand for sarcolectin with survival of patientswith bronchial carcinoma. Eur J Cancer 30A:653–657.

Kayser, K., J. Bubenzer, G. Kayser, S. Eichhorn,T.V. Zemlyanukhina, N.V. Bovin, S. André, J.Koopmann, H.-J. Gabius (1995) Expression oflectin-, interleukin-2-, and histo-blood group-binding sites in prostate cancer and its correla-tion to integrated optical and syntactic structureanalysis. Anal Quant Cytol Histol 17: 135–142.

Kayser, K., H.-J. Gabius (1997) Graph theory andthe entropy concept in histochemistry. ProgHistochem Cytochem 32/2: 1–106.

Kayser, K., M. Heil, H.-J. Gabius (1989) Is the pro-file of binding of a panel of neoglycoproteinsuseful as diagnostic marker in human lung can-cer? Pathol Res Pract 184: 621–629.

Kopitz, J., C. von Reitzenstein, M. Burchert, M.Cantz, H.-J. Gabius (1998) Galectin-1 is amajor receptor for ganglioside GM1, a productof the growth-controlling activity of a cell sur-face ganglioside sialidase, on human neuro-blastoma cells in culture. J Biol Chem 273:11205–11211.

Laine, R.A. (1997) The information-storing poten-tial of he sugar code; in Gabius H-J, S Gabius(eds): Glycosciences: Status and Perspectives.Weinheim, Chapman & Hall, pp 1–14.

Laurila, P., J. Virtanen, J. Wartiovaara, S. Stenman(1978) Fluorescent antibodies and lectins stainintracellular structures in fixed cells treatedwith nonionic detergent. J Histochem Cyto-chem 26: 251–257.

Lee, R.T., H.-J. Gabius, Y.C. Lee (1992) Ligandbinding characteristics of the major misteletoelectin. J Biol Chem 267: 23722–23727.

Lee, R.T., H.-J. Gabius, Y.C. Lee (1994) The sugar-combining area of galactose-specific toxic lectinof mistletoe extends beyond the terminal sugarresidue: Comparison with a homologous toxiclectin, ricin. Carbohydr Res 254: 269–276.

Lee, R.T., Y.C. Lee (1997) Neoglycoconjugates;in Gabius H-J, S Gabius (eds): Glycosciences:Status and Perspectives. Weinheim, Chapman& Hall, pp 55–77.

Lee, Y. (1987) Lectin reactivity in human largebowel. Pathology 19: 397–401.

Lee, YC, Lee RT (eds) (1994) Neoglycoconju-gates. Preparation and Applications. Orlando,Academic Press.

Lee, Y.S. (1988) Lectin expression in neoplasticand non-neoplastic lesions of the rectum.Pathology 20: 157–165.

Mann, P.L., R.E. Waterman (1998) Glycocoding asan information management system in embry-onic development. Acta Anat 161: 153–161.

McGarrity, T.J., L.P. Pfeifer, A.B. Abt (1989)Lectin histochemistry of adenomatous polyps.Not a predictor of metachronous lesions. Can-cer 64: 170–1713.

Neutra, M.R., T.L. Philips, E.L. Mayer, D.J.Fishkind (1987) Transport of membrane-bound macromolecules by M cells in follicle-associated epithelium of rabbit Peyer’s patch.Cell Tissue Res 247: 537–546.

Ofek, I., N. Sharon (1988) Lectinophagocytosis: Amolecular mechanism of recognition betweencell surface sugars and lectins in the phagocy-tosis of bacteria. Infect Immunol 56: 539–547.

Ohannesian, D.W., D. Lotan, P. Thomas, J.M. Jes-sup, M. Fukuda, H.-J. Gabius, R. Lotan (1995)Carcinoembryonic antigen and other glyco-conjugates act as ligands for galectin-3 inhuman colon carcinoma cells. Cancer Res 55:2191–2199.

Orntoft, T.F., N.C. Langkilde, H. Wiener, P.D.Ottosen (1991) Cellular localization of PNAbinding in colorectal adenomas: Compari-son with differentiation, nuclear: Cell heightRatio and effect of desialylation. APMIS 99:274–281.

Ota H., J. Nakayama, T. Katsuyama, M. Kanai(1988) Histochemical comparison of specific-ity of three bowel carcinoma-reactive lectins,Griffonia simplicifolia agglutinin, peanut ag-glutinin and Ulex europaeus agglutinin-I. ActaPathol Jpn 38: 1547–1549.

Petrie, W.A., R.D. Smith, P.H. Schlesinger, C.F.Murphy, J.L. Ravidin (1987) Isolation ofthe galactose-binding lectin that mediates thein vitro adherence of Entamoeba histolytica.J Clin Invest 80: 1238–1244.

Plendl, J., F. Sinowatz (1998) Glycobiology of theolfactory system. Acta Anat 161: 234–253.

Podolsky, D.K. (1989) The colonic goblet cell andglycoprotein heterogeneity. Immunol Invest18: 485–497.

Radzun, H.J., M.R. Parwaresch, A.C. Feller,M.L. Hansmann (1984) Monocyte/macro-phage-specific monoclonal antibody Ki-M1recognizes interdigitating reticulum cells. AmJ Pathol 117: 441–450.

Rhodes, J.M., R.R. Black, A. Savage (1986)Glycoprotein abnormalities in colonic carci-nomata, adenomata, and hyperplastic polypsshown by lectin peroxidase histochemistry.J Clin Pathol 39: 1331–1334.

Rüdiger, H. (1998) Plant lectins – more than justtools for glycoscientists. Occurrence, structureand possible functions of plant lectins. ActaAnat 161: 130–152.

Sata, T., J. Roth, C. Zuber, B. Stamm, S.J. Rinderle,I.J. Goldstein, P.U. Heitz (1992) Studies on theThomsen-Friedenreich antigen in human colonwith the lectin Amaranthin. Normal and neo-plastic epithelium express only cryptic T anti-gen. Lab Invest 66: 175–186.

Schneller, M., S. André, J. Cihak, H. Kaltner, H.Merkle, G.J. Rademaker, J. Haverkamp, J.Thomas-Oates, U. Lösch, H.-J. Gabius (1995)Differential binding of two chicken β-galac-toside-binding lectins to homologous lym-phocyte subpopulations and evidence forinhibitory activity of the dimeric lectin on stim-ulated T cells. Cell Immunol 166: 35–43.

Schoeppner, H.L., A. Raz, S.B. Ho, R.S. Bresalier(1994) Expression of an endogenous gal-actose-binding lectin correlates with neoplas-tic progression in the colon. Cancer 75:2818–2826.

Sharma, A., R.A. DiCioccio, H.J. Allen (1992)Identification and synthesis of a novel 15 kDaβ-galactoside-binding lectin in human leuko-cytes. Glycobiology 2: 285–292.

Sharon, N. (1987) Bacterial lectins, cell-ceelrecognition and infectious disease. FEBS Lett217: 145–157.

Sharon, N. (1998) Glycoproteins now and then.Acta Anat 161: 7–17.

Smetana, K., Jr., Slavik, E. Vancova, I. Fischer,F.-T. Liu, M. Burchert, X. Dong, H.-J. Gabius(1998) Fusion of macrophages on an implantsurface is associated with down-regulated ex-pression of ligands for galectin-1 and -3 in therat. Biomaterials, in press.

Solís, D., A. Romero, H. Kaltner, H.-J. Gabius, T.Díaz-Mauriño (1996) Different architecture ofthe combining site of the two chicken galectinsrevealed by chemical mapping studies withsynthetic ligand derivatives. J Biol Chem 271:12744–12748.

Varki, A. (1994) Selectin ligands. Proc Natl AcadSci USA 91: 7390–7397.

Villalobo, A., H.-J. Gabius (1998) Signaling path-ways for transduction of the initial messageof the glycocode into cellular response. ActaAnat 161: 110–129.

von der Lieth, C.-W., H.-C. Siebert, T. Kozár, M.Burchert, M. Frank, M. Gilleron, H. Kaltner,G. Kayser, E. Tajkhorshid, N.V. Bovin, J.F.G.Vliegenhart, H.-J. Gabius (1998) Lectin Li-gands: New insights into their conformationsand their dynamic behavior and the discoveryof conformer selection by lectins. Acta Anat161: 91–109.

von Rosen, L., B. Podjaski, I. Bettman, H.F. Ott(1981) Observations on the ultrastructure andfunction of the so-called ‘microfold’ or ‘mem-branous’ cells (M cells) by means of per-oxidase as a tracer. An experimental studywith special attention to the physiologicalparameters of resorption. Virchows Arch 390:289–312.

Vyakarnam, A., S.F. Dagher, J.L. Wang, R.J. Pat-terson (1997) Evidence for a role for galectin-1in pre-mRNA splicing. Mol Cell Biol 17:4730–4737.

Wasano, K., Y. Hirakawa (1997) Recombinantgalectin-1 recognizes mucin and epithelial cellsurface glycocalyces of gastrointestinal tract.J Histochem Cytochem 45: 275–283.

Wolf, J.L., W.A. Bye (1984) The membranousepithelial (M) cell and the mucosal immunesystem. Annu Rev Med 35: 95–112.

Yoshioka, H., M. Inada, K. Ogawa, G. Ohshio,H. Yamabe, Y. Hamashima, T. Miyake (1989)Lectin histochemistry in ulcerative colitis andCrohn disease. J Exp Pathol 4: 69–78.

Zanetta, J.-P. (1998) Structure and functions oflectins in the central and peripheral nervoussystem. Acta Anat 161: 180–195.

Zschäbitz, A. (1998) Glycoconjugate expressionand cartilage development in the cranial skele-ton. Acta Anat 161: 254–274.


Acta Anat 1998;161:219–233

Dow

nloa

ded

by:

UB

der

LM

U M

ünch

en

129.

187.

254.

47 -

8/1

2/20

13 1

2:22

:45

PM

u. brinck detection of inﬂammation- and a neoplasia ... · oooooooooooooooooooo detection of...

Documents